JP3219988B2 - Disk device having security function and password management method in the device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a disk device having a security function and a password management method in the disk device.

[0002]

2. Description of the Related Art In recent years, a disk device having a security function for protecting (protecting) data recorded on a disk as a recording medium from being leaked to a third party or tampered with by a third party has been developed. Has appeared.

To protect (protect) data recorded on a disk of this type of disk device, a user performs an operation for setting a password in the disk device from the host device. Then, the password specified by the user is stored in the BIOS (Basic Input / Output) of the host device.
System) and input to the disk unit.

[0004] Here, there have been conventionally known two password setting methods in a disk device. First
In the password setting method, a user-specified password transferred from the host device is set and stored in an electrically rewritable nonvolatile memory of the disk device, for example, an EEPROM.

In the second password setting method, a user-specified password transferred from the host device is set and stored at a specific position on a disk of the disk device. In either method, if a hard reset (initialization by a command from the host device) or a power-on reset (initialization by turning on the power) of the disk device is performed after setting the password, the disk device is locked. In other words, unless the same password as the set password is input by the user, reading / writing (for the disk) on the disk device is prohibited, and data recorded on the disk is protected.

[0006]

However, a password is set (saved) in a conventional disk device to which the above-described first password setting method is applied, that is, in a rewritable nonvolatile memory typified by an EEPROM. In such a conventional disk device, the EEPROM is usually used.
(Rewritable non-volatile memory represented by) is mounted on a printed wiring board (PCB). Therefore, when the printed wiring board is replaced, no password is set, and There is a problem that data access becomes possible.

On the other hand, in the conventional disk device to which the above-mentioned second password setting method is applied, the password is set (stored) on the disk, so that there is no such a fear.

In a disk device having a security function, if a set password is lost due to data destruction or the like, even if a correct password is input, data access on the disk device becomes impossible.

For this reason, in a conventional disk device to which the above-described second password setting method is applied, that is, a conventional disk device in which a password is set (stored) at one specific position on a disk, the password is set to If the data in the set area is destroyed for some reason, there is a problem that the device becomes fatal. On the other hand, the disk device to which the first password setting method is applied does not have such a fear.

Here, the causes of data destruction on the disk are listed below. (1) A defect occurs in a sector in an area where a password is written, and the sector becomes a late defect.

(2) The power is turned off while data (password) is being written. (3) HDI (Head Disk) where the head collides with the disk
Interference, i.e., a head crash, destroys the entire track or disk surface at the point of collision.

If any of the above (1) to (3) occurs in the area at the specific position, the password is lost, and the data of the apparatus cannot be recovered. The present invention has been made in consideration of the above circumstances, and has as its object to provide a password information setting area (password information area) secured on a disk.
Another object of the present invention is to provide a disk device having a security function capable of backing up data by another password information area even if the data is destroyed for some reason, and a password management method in the device.

Another object of the present invention is to allow a password information area secured on another disk surface to be backed up even if the entire disk surface in which the password information area is secured becomes inaccessible due to a head crash or the like. Is to do.

It is still another object of the present invention to enable password information to be accessed efficiently. Still another object of the present invention is to improve the integrity of password information.

[0015]

SUMMARY OF THE INVENTION The present invention provides at least one
In a disk device having a plurality of disks and having a security function of permitting disk access only when a password matching the set password is given from the host device, the password information is recorded.
Multiple password information areas for the entire disk
That is, a structure in which the disks are dispersedly arranged on at least two disk surfaces.
And read error from all password information areas.
Area with a low probability of occurrence of at least two disk surfaces
Predetermined number of effective password information areas to be distributed
Select the password and enter it in this predetermined number of valid password information areas.
And wherein the configuration and was able to write the same password information Re respectively.

In such a configuration, a plurality of disks are provided.
Read from all password information areas distributed on the desktop
Areas with low error probability are distributed over multiple disk surfaces
Selected as the valid password information area
The same path for each of the selected valid password information areas
Word information is written. Thus, the same password
Password information is distributed on multiple disk surfaces.
Since it is stored in the host information area, that is,
Backup candidates that can be compared with passwords
Can be held on multiple disk surfaces,
Due to an accident such as a late defect or power interruption, a certain
Even if the information in the password information area is lost, other valid paths
Password given from host device from word information area
Password information required for comparison with the password.
You. In addition, the effective password information area is
Due to the distributed arrangement, head crash etc. occur,
Even if the entire disk surface becomes inaccessible,
Valid password information area located on another disk
Password from the host device from the password
Get necessary password information and backup
High reliability. In addition, the valid password information area is
Multiple password information distributed on multiple disk surfaces
Area where the read error occurrence probability selected from the
Area, so when reading the valid password information area
It is possible to avoid the occurrence of retry processing, etc.
You can access password information in the process.

[0017]

[0018]

[0019]

[0020]

Here, in order to select a password information area having a low probability of occurrence of a read error, test data or password information is written in all the password information areas, and then each area is read a plurality of times to be readable. A method of selecting a predetermined number of regions having a large number of times in descending order is applicable. However, in a disk device having an ECC correction function for read data (read sector data) (such a disk device is generally used),
It is more effective to turn off the ECC collection function and perform the above read. In addition, a reference value of the number of times that data can be read is provided, and if the number of areas equal to or larger than the reference value is less than a predetermined number, a disk device error may be determined.

In addition, as a method of selecting a password information area with a low probability of occurrence of a read error, test data or password information is written in all the password information areas, and then each area is read and read normally. A method of selecting a predetermined number of regions in which the number of retries is equal to or less than a certain value from the region in an ascending order of the number of retries is also applicable.

Here, the selected valid password information
Information area, that is, effective
The position information of the password information area is stored in a non-volatile memory
Access to password information by recording on
The access destination is recorded in the recording means.
It is good to decide based on the location information. In addition, the effective
Information indicating the access order to the password information area
Recorded in the recording means and access the password information
In this case, the access order is stored in the
It may be determined based on the access order information. In particular,
The access order information indicates the order indicated by the selected valid path.
Word information areas are sorted in ascending order of read error probability.
It is good to match. In addition, the above-mentioned order is
Password information areas in the order that they can be accessed in the shortest time
They may be matched.

By doing so, the effective password information area is accessed in the order of the lowest error occurrence probability or in the order in which the access speed is prioritized, so that the password information reading without waste is realized. In particular, by writing the password information in the order indicated by the order information, that is, by writing the password information (password setting) and reading the password information in the same order, power-off or the like during password setting can be prevented. As a countermeasure in case of occurrence, it becomes possible to enhance the integrity of the password information.

[0025]

[0026]

[0027]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing a configuration of a magnetic disk drive according to one embodiment of the present invention.

The magnetic disk device shown in FIG. 1 has, for example, three disks 1-0 to 1-2 as media on which data is recorded. These discs 1-0 to 1-2
Corresponding to each disk surface (data surface), a head (magnetic head) 2 used for data writing (data recording) on the disk surface and data reading (data reproduction) from the disk surface is provided. I have.

A large number of concentric tracks are formed on both surfaces of the disk 1-i (i = 0 to 2).
Used for positioning control, etc. (including cylinder data indicating cylinder number and burst data for indicating position error in cylinder indicated by cylinder data by waveform amplitude)
A plurality of servo areas on which servo data is recorded are arranged at equal intervals. These servo areas are on disk 1-i
Above, they are arranged radially across each track from the center. The area between the servo areas is a user area. One servo area is followed by one user area to form one servo sector. A plurality of data sectors are set in the user area of each servo sector.

The disk 1-i has a spindle motor (SP)
M) High speed rotation due to 3. The head 2 is a carriage 4
The carriage 4 moves in the radial direction of the disk 1-i. The carriage 4 includes a voice coil motor (VCM) 5
Driven by

SPM (spindle motor) 3 and VCM
(Voice coil motor) 5 is connected to motor driver 6. Motor driver 6, another for driving the SPM 3 by supplying a control current to SPM 3, and drives the VCM 15 by supplying a control current to VCM 15. The value of the control current (control amount) is controlled by a CPU (microprocessor) 10.
, And given as a digital value, for example.

Each head 2 is connected to a head IC 7 mounted on a flexible printed circuit (FPC). The head IC 7 controls switching of the head 2 and inputs and outputs read / write signals to and from the head 2, and has a head amplifier 71 that amplifies an analog output read by the head 2.

The head IC 7 is connected to a read / write IC 8 (read IC, read / write circuit) 8.
The read / write IC 8 roughly has an encode / decode function for processing user data and a signal processing function for processing servo data.

The read / write IC 8 receives an analog output (read signal of the head 2) read from the disk 1-i by the head 2 and amplified by the head amplifier 71 in the head IC 7, and is required for a data reproducing operation. Signal processing,
For example, signal processing for converting an analog output into NRZ data and transferring the data to a disk controller (HDC) 14 is performed by a decoding function. Read / write IC8
Is a signal processing required for the data recording operation, for example, HDC1
4 modulates the NRZ data (write data) sent from disk 4 and writes it to disk 1-i (eg, 2-7,
1-7 modulated data) and sends the signal to the head IC 7 by the encode function.

The read / write IC 8 also executes, by a signal processing function, a servo data reproduction process required for servo processing such as head positioning control, in addition to the above-described normal user data recording / reproduction process. That is, read / write I
C8 processes the servo data in the servo area read by the head 2 and outputs a data pulse including cylinder data to the servo processing circuit 9. Also read / write I
C8 samples and holds the peak of the burst data (in the servo data) and outputs it to the servo processing circuit 9.

The servo processing circuit 9 includes a read / write IC
In response to the data pulse and the burst data from the control unit 8, signal processing necessary for servo processing is executed. That is, the servo processing circuit 9 has a decoding function for extracting and decoding cylinder data (cylinder number) and the like from the data pulse from the read / write IC 8, and a timing generation function such as a write gate. Also, the servo processing circuit 9 has a read / write I
A / D burst data (analog signal) from C8
It also has an A / D conversion function of converting (analog / digital) and outputting it to the CPU 10.

The CPU 10 is, for example, a one-chip microprocessor. The CPU 10 controls each unit in the magnetic disk device according to a control program stored in the ROM 11. In the control by the CPU 10, the head 2 is moved to a target position (by controlling the driving of the VCM 5 via the motor driver 6) in accordance with the servo data (the middle cylinder data and the burst data) extracted by the servo processing circuit 9. Control, read / write data transfer control by controlling the HDC 14, and password management control directly related to the present invention.

The CPU 10 has a ROM 1 as a non-volatile memory in which a control program (firmware) for controlling each unit in the magnetic disk device is stored.
1, a RAM 12 as a rewritable memory that provides a work area of the CPU 10, a storage area for parameters used by the CPU 10, and the like, and a rewritable nonvolatile memory that stores parameters for controlling the magnetic disk device. EEPROM 13 and a disk controller (HDC) 14 are connected.

The disk controller (HDC) 14
It controls communication of commands and data with a host device (not shown), and controls data communication with the read / write IC 8 (via the disk 1-i). The HDC 14 is connected to a buffer memory 15 that stores read / write data in a cache system, such as a RAM, for example, and a host interface 16. The host interface 16 serves as an interface between the HDC 14 and the host device.
Performs command and data communication with the host device via the host interface 16.

Now, each disk 1-i (i = 0 to 0) in FIG.
2), as shown in FIG.
The system area 101 is allocated to a predetermined area of −0, 100-1, for example, a predetermined plurality of continuous tracks. The system area 101 is set in a different area from the data area, and is used for storing information on a defect on a corresponding disk surface. For example, four password information areas SSD (Security Save Data) 0 to SSD3 used for storing password information are secured in a predetermined track 102 of the system area 101. Here, the password information area SSD
0 to SSD3 are distributed on the track 102,
The position is, for convenience, on both disk surfaces 10 of each disk 1-i.
It is assumed that 0-0 and 100-1 are the same (however, they need not be the same). The disk surface 1 of the disk 1-0
The head numbers of 00-0 and 100-1 are “0” and “1” (that is, heads # 0 and # 1), and the disk surface 100 of the disk 1-1
-0 and 100-1 are head numbers "2" and "3" (that is, heads # 2 and # 3), and disk surfaces 100-0 and
The head numbers of 100-1 are "4" and "5" (that is, head #
4, # 5).

Each password information area SSDj (j = 0 to
3) is composed of, for example, two consecutive sectors (data sectors) Sn and Sn + 1 as shown in FIG. Each of the sectors Sn and Sn + 1 includes an ID section, a data section, and an ECC for error detection and correction. The password information is distributed and stored in both data portions of the sectors Sn and Sn + 1. The password information includes a header section 41 indicating password information (security information), a flag 42 indicating the presence or absence of a password setting, a counter section 43 indicating the version of the password information (initial value 1), and a password 44. , And a checksum 45 (of the information) for detecting an error in the password information (across the sectors Sn and Sn + 1).

Next, the password management control in this embodiment will be described. First, in the present embodiment, one of the first to third password management control methods is applied. Therefore, the operation when the first to third password management control methods are applied will be described in order.

(1) Operation When the First Password Management Control Method is Applied First, the operation when the first password management control method is applied will be described with reference to the flowcharts of FIGS.

In the first password management control method, FIG.
In the manufacture of the magnetic disk device, the valid SSD selection / registration processing according to the flowchart of FIG. 5 is performed. The valid SSD selection / registration process is started when a host device connected to the host interface 16 of the magnetic disk device of FIG. 1 sends a start command of the valid SSD selection / registration process to the magnetic disk device. You.

The valid SSD selection / determination processing start command from the host device is transmitted to the host interface 16 and the HD.
It is passed to the CPU 10 via C14. CPU 10
When a valid SSD selection / registration processing start command is received from the host device, the system area 1 secured on each of the disk surfaces 100-0 and 100-1 of the three disks 1-0 to 1-2.
01 is written in all of the password information areas SSD0 to SSD3 assigned to the storage area 01, ie, a total of 24 pieces of password information areas SSD. Is performed (step S1).

Next, the CPU 10 executes the above-mentioned discs 1-0 to 1
-2 reads the test data written in the password information areas SSD0 to SSD3 secured on the disk surfaces 100-0 and 100-1, respectively, and selects (detects) the password information area SSD with a low error probability. A selection process is performed (step S2).

In this step S2, as a method of selecting an area SSD having a low error probability, the operation of the ECC collection is turned off for each SSD, and each SSD is read a plurality of times (predetermined number of times) to be readable. A method of selecting from those having a large number of times, or a method of setting the number of read retries and selecting an area SSD that can be read with a certain number of retries or less can be applied.

Here, the region SS where the error probability is low is
As D, three areas (SSD1, SSD2, SSD3) from the disk surface 100-0 (head # 0) of the disk 1-0,
One area (SSD2) from the disk surface 100-1 (head # 1) of the disk 1-0, the disk surface 100 of the disk 1-1
-0 (head # 2) to two areas (SSD1, SSD3),
Two areas (SSD0, SSD2) from the disk surface 100-0 (head # 4) of the disk 1-2, and one area (SSD1) from the disk surface 100-1 (head # 5) of the disk 1-2.
) Are selected (detected).

When the CPU 10 selects (detects) an area SSD having a low error probability from the 24 password information areas SSD in step S2, the predetermined number N (N is an integer of 2 or more; Where N =
The SSDs of 4), which are distributed over a plurality of (here, N) disk surfaces, are selected as valid SSDs (step 3). Therefore, in the present embodiment in which the nine areas SSD are selected in step S2, for example, from the disk surface 100-0 (head # 0) of the disk 1-0 (the SSDs 1 and S2 selected in step S2).
(The lowest error probability among SD2 and SSD3)
One area SSD1 (which may be SSD2 or SSD3) is located on the other disk surface 10 of the disk 1-1.
One area SSD2 from 0-1 (head # 1) has the highest error probability among the SSD1 and SSD3 selected in step S2 from the disk surface 100-0 (head # 2) of the next disk 1-1. (Low) one area SSD1 (or SSD3 may be used), and (because there is no area SSD with a low error probability on the disk surface 100-1 of the disk 1-1), and the disk surface of the next disk 1-2 100-0
(Head # 4) to (SSD selected in step S2)
One area SSD0 (or SSD2 may be used) which has the lowest error probability among 0 and SSD2.

In this embodiment, if the number of valid SSDs selected (detected) in step S2 is less than N (= 4), or if the number of valid SSDs is N or more, a plurality (in this case, If they are not distributed over (N or more) disk surfaces, the corresponding magnetic disk device is treated as defective.

When selecting the N valid SSDs (valid password information areas SSD) in step S3, the CPU 10 uses the information having the position information of the N valid SSDs as information on the valid SSDs (valid SSD information) as EEP.
The valid SSD is registered in a predetermined area of the ROM 13 and the valid SSD selection / registration processing is terminated (step S4).

Next, a process (lock process, password) when the magnetic disk device in which the information on the N valid SSDs (valid SSD information) is registered in the EEPROM 13 is shipped by the user after shipment is performed. The setting process will be described with reference to the flowchart of FIG. However, after the valid SSD selection / registration process,
Until shipment, the counter section 43 of the N valid SSDs selected in the valid SSD selection / registration processing (see FIG. 3)
Is initialized to a value of 1. If the counter unit 43 having a value of 1 is provided in the test data, this initial setting process is unnecessary.

First, a user performs an operation for setting a desired password on the magnetic disk device of FIG. 1 and locking the disk device from the host device. Thus, a password setting command (set password command) including a user-specified password is sent from the host device to
To the magnetic disk device.

The CPU 10 receives the set password command sent from the host device via the host interface 16 and the HDC 14 (step S11). Then, the CPU 10 sets information on the valid SSD registered in a predetermined area of the EEPROM 13 (valid SSD information).
(Step S12), and based on the valid SSD position information indicated by the valid SSD information, the disks 1-0 to 1-2 are read.
A total of N valid SSDs (valid password information areas S) distributed on N surfaces (here, N = 4) of the respective disk surfaces 100-0 and 100-1 (that is, a total of six disk surfaces).
Then, a disk read process for sequentially reading information from SD) is performed (step S13).

Next, every time the CPU 10 reads the information of one valid SSD, the CPU 10 increments the content of the counter 43 (see FIG. 3) in the read information by one, and the value after the increment (the first lock processing). In this case, “2”) is stored in the counter unit 43, a password specified by the user is stored as the password 44, and the password information in the format of FIG. A disk write process is performed (step S14). The password information is provided with a checksum 45, and when the password information is dispersedly written in the two sectors Sn and Sn + 1 constituting the valid SSD, for example, the power is shut off after writing to the sector Sn. Therefore, if the rest of the password information is not written in the sector Sn + 1, the checksum becomes inconsistent even if there is no ECC error in the subsequent read, and
The error of the valid SSD can be detected.

The CPU 10 executes the above-described steps S12 to S12.
14 is repeatedly executed for the N valid SSDs (step S15), and the lock processing ends. When a hard reset or power-on reset of the magnetic disk device is performed after the lock processing, the magnetic disk device is locked, and the same password as the set password is not input by the user (the locked state is not released). As long as (disks 1-0 to 1
Read / write (for -2) is prohibited, disc 1-0
The data recorded in 1-2 are protected.

Next, the lock release processing will be described with reference to the flowchart of FIG. First, when using the magnetic disk device of FIG. 1, the user performs an operation for releasing the locked state of the magnetic disk device from the host device. Thus, a command for unlocking (unlock command) including the password specified by the user is sent from the host device to the magnetic disk device of FIG.

The CPU 10 transmits the unlock command sent from the host device to the host interface 16 and the HD.
It is received via C14 (step S21). Then C
The PU 10 refers to information (valid SSD information) regarding the valid SSD registered in a predetermined area of the EEPROM 13 (step S22), and based on the valid SSD position information indicated by the valid SSD information, discs 1-0 to A total of N valid SSDs (valid password information) distributed on N surfaces (N = 4 in this case) among the 1-2 disk surfaces 100-0 and 100-1 (ie, a total of 6 disk surfaces) Area SSD)
Then, a disk read process for sequentially reading the setting information (password information) is performed with the ECC collection operation ON (ON) (step S23).

Each time the CPU 10 reads the password information from one valid SDD, the CPU 10 determines whether the reading is successful and whether the value of the read password information matches the checksum 45 added to the password information. Is checked (step S24), and if so (when the checksum is matched), the password information is temporarily stored (step S25).

The CPU 10 executes the above steps S22 to S
25 is repeatedly executed for the N valid SSDs (step S26), and among all the pieces of password information stored at that time, the password information having the largest value of the counter unit 43 is correctly set to the password. One is selected as the latest password information (step S27).

Next, the CPU 10 compares the password 44 in the selected password information with the password designated by the user (the password attached to the unlock command sent from the host device) (step S28). Only if it is (step S29), the locked state is released (step S30).

Here, the reason why the password information which does not match the checksum even if it can be read correctly is excluded from the password information selection processing in step 27 is as follows.

First, when the password information is distributed and written to the two sectors Sn and Sn + 1 constituting the valid SSD, if the power supply is cut off after writing to the sector Sn, the data section of the sector Sn is updated to the latest data. Although the first half of the password information is written, the latter half of the previous password information remains in the data portion of the sector Sn + 1. In this case, the absence of a read error from the sectors Sn and Sn + 1 does not mean that the read password information is correct. Therefore, in the present embodiment, the checksum 45 is added to the password information, and even if there is no read error, if the read password information becomes inconsistent with the checksum, the password information is turned off at the time of writing (or (The lock process failed), and the password information selection processing in step 27 is performed as described above, assuming that the first half and the second half of the password information are not written at the same time. Outside.

The reason why the password information having the largest value of the counter 43 in the process of step 27 is selected as a target for password comparison is as follows. First,
When the power is turned off after writing the password information to one of the N valid SSDs, the previous password information remains in the remaining valid SSDs. In this case, even if there is no read error and there is no checksum mismatch, the read password information cannot be made correct (latest one). Therefore, in the present embodiment, the password information is provided with the counter 43, and when newly setting the password information, the content of the counter 43 is incremented by 1 and reset.
The latest version of the password information is indicated. Therefore, as the latest password information to be subject to the lock release determination, of the password information having no read error and no checksum inconsistency,
By selecting the password information having the maximum value of the counter section 43, correct comparison with a user-specified password is guaranteed even if the power-off or lock processing during writing of password information fails in the middle.

In addition, in this embodiment, a plurality of valid SSs
D (here, four) are distributed on each disk surface. Therefore, even if data of some valid SSDs is destroyed due to the factors described in the section of [Prior Art], other valid SSDs are destroyed. S
If the password information of the SD is available and thus the correct password is given, the lock state is released.

(2) Operation when the Second Password Management Control Method is Applied According to the first password management control method described above, the password 44 in the password information affected by the power cutoff at the time of writing the password information or the like is used. Has been described as having the counter section 43 in the password information so that the password information is not compared with the user-specified password.
The counter unit 43 is not always necessary. Therefore, an operation in the case where the second password management control method in which the counter 43 in the password information is unnecessary is applied will be described.

In the second password management control method, the process according to the flowcharts of FIGS. 7 to 9 is performed as follows instead of the flowcharts of FIGS. 4 to 6 in order to eliminate the need for the counter unit 43 in the password information. To do. In FIGS. 7 to 9, the same processing parts as those in FIGS. 4 to 6 are denoted by the same reference numerals.

First, in the valid SSD selection / registration processing, the processing in step S4a in FIG. 7 is performed instead of the processing in step S4 in FIG. Here, the information having the position information of the N valid SSDs and the information of the order of access (priority order) is stored in the information (valid SSDs) relating to the N valid SSDs.
(Information) in the EEPROM 13. This order information is generated based on, for example, the read error occurrence probabilities of the N valid SSDs, and the lower the read error occurrence probability, the higher the rank. In addition, when sequentially accessing the N valid SSDs, the valid SSDs that can be accessed (seeked) in the shortest time can be used (an order in which access speed is prioritized).

Next, lock processing (password setting processing)
Then, instead of steps S12 to S15 in FIG.
The processing of steps S31 to S35 is performed. here,
Valid SS registered in a predetermined area of the EEPROM 13
With reference to the position information and the order information in the D information (step S31), first, password information including the user-specified password is written to the first (highest priority) valid SSD (step S32). Next, the position information and the order information in the valid SSD information are referred to again (step S33), and password information including the user-specified password is written to the next valid SSD in the order (step S34). Then, the above steps S33 and S34 are repeatedly executed for the (N-1) valid SSDs excluding the first valid SSD (step S35), and the lock processing ends.

Next, in the unlocking process, step S4 in FIG. 9 is performed instead of steps S22 to S28 in FIG.
1 to S47 are performed. Here, the EEPROM 13
The position information and the order information in the valid SSD information registered in the predetermined area are referred to (step S41), and first, the password information is read from the first (highest priority) valid SSD in the order (step S42). . Next, it is checked whether the read is successful and whether the value of the read password information matches the checksum 45 attached to the password information (step S43).

If the read fails or if the read succeeds but the checksum is not consistent, if the password information is not read from the last valid SSD (step S44), the password is not read from the valid SSD information. The position information and the order information are referred to again (step S45), and the password information is read from the next effective SSD (step S46). Then, the process returns to step S43.

On the other hand, if it is determined in step S43 that the reading of the password information has succeeded and the checksums are consistent, the password 44 in the password information and the password designated by the user (transmitted from the host device). (Step S47), and only when the passwords match (Step S29), the locked state is released (Step S30).

As described above, by reading the valid SSDs in the same order as the writing order of the password information, even if the password information is not provided with the counter section (43) as described above, the password information can be written at the time of writing the password information. Password information can be preserved as a countermeasure in the event that the power-off or lock processing has failed.

In the above description, the valid SSD information is provided with order information, and the valid SSD information is stored in the order indicated by this information.
Although the password information for D is read / written, the password information for the valid SSD is determined in accordance with the order of the position information (position information indicating the positions of the N valid SSDs) in the valid SSD information. May be performed. In this case, the valid SSD
The information in the order of the information can be made unnecessary. However,
In the case of using the order information, the order is minimized by matching the order with the order of the lowest probability of occurrence of a read error or the order in which the effective SSD can be accessed in the shortest time (the order in which access speed is prioritized). You can read password information in time.

(3) Operation when the Third Password Management Control Method is Applied In the first and second password management control methods described above, the N valid SSDs are fixed when the magnetic disk device is shipped. Have been. Therefore, if the data of some of the valid SSDs is destroyed for some reason after shipment, the number of actually usable valid SSDs decreases.
Even in this case, backup can be performed using the remaining valid SSDs. However, there is a problem when the number of remaining valid SSDs becomes extremely small (although the probability that such a case will occur is almost equal to 0).

The operation in the case where the third password management control method in which the valid SSD is not fixed will be described with reference to the flowcharts of FIGS. 10 and 11. Note that, in the third password management control method, similarly to the second password management control method, the password information does not include the counter unit 43.

In this case, the valid SSD selection / registration processing as shown in the flowchart of FIG. 4 (or FIG. 7) at the time of manufacturing the magnetic disk drive of FIG. 1 is not performed, and the disks 1-0 to 1-2 are simply processed. Only four password information areas SSD0 to SSD3 (a total of 24 password information areas SSD) are allocated in the system area 101 secured on each of the disk surfaces 100-0 and 100-1.

When locking the magnetic disk drive of FIG. 1, the user performs an operation for that from the host device.
As a result, a password setting command (set password command) including a user-specified password is sent from the host device to the magnetic disk device of FIG. 1, and the lock process (password setting process) described below is performed according to the flowchart of FIG. Will be

First, the CPU 10 sends the set password command sent from the host device to the host interface 1.
6 via the HDC 14 (step S5).
1). Then, the CPU 10 sets a flag indicating that a lock process is being performed in a predetermined area of the EEPROM 13 (step S52).

Next, the CPU 10 stores the password information areas SSD0 to SSD3 allocated in the system area 101 secured on the disk surfaces 100-0 and 100-1 of the disks 1-0 to 1-2, respectively. Password information having a password designated by the user as the password 44 and indicating that the flag 42 is set with the password 42 in the format of FIG. 3 (but not having the counter 43) is written (step S53).

Next, the CPU 10 executes the above-mentioned discs 1-0 to 1
-2 reads the password information written in the password information areas SSD0 to SSD3 secured on the respective disk surfaces 100-0 and 100-1 and performs an SSD selection process for selecting (detecting) an SSD having a low error probability. (Step S5
4).

In step S54, as in the method of step S2, the operation of the ECC collection is turned off (O
N) and a method of selecting from among those that can be read, or a method of setting the number of read retries and selecting an area SSD that can be read with a certain number of retries or less can be applied. Here, the disk surface 100-0 of the disk 1-0 (the head #
0) to 3 areas (SSD1, SSD2, SSD3), and 1 from the disk surface 100-1 (head # 1) of the disk 1-0.
Area (SSD2), disk surface 100-0 of disk 1-1
(Head # 2), two areas (SSD1, SSD3), and two areas from disk surface 100-0 (head # 4) of disk 1-2.
Area (SSD0, SSD2) and one area (SSD1) from the disk surface 100-1 (head # 5) of the disk 1-2.
It is assumed that a total of nine areas have been selected (detected).

The CPU 10 determines at step S54 that
When an area SSD having a low error probability is selected (detected) from the plurality of password information areas SSD, a predetermined number N (here, N = 4) of SSDs is selected from the plurality, and a plurality (here, N = 4) ) Are selected as valid SSDs (step 55). Therefore, in the present embodiment in which the above-mentioned nine areas SSD are selected in step S54, for example, from the disk surface 100-0 (head # 0) of the disk 1-0 (the SSD1, SSD2,. SSD
1 among the three areas SSD1 with the lowest probability of error)
Is the other disk surface 100-1 of the disk 1-1.
One area SSD2 from (head # 1) is the next disk 1
-1 from the disk surface 100-0 (head # 2) (one of the SSD1 and SSD3 selected in step S54, which has the lowest error probability), and one area SSD1 (the disk surface 10 of the disk 1-1). -1 because there is no area SSD with a low probability of error)
One area SSD0 is selected from the disk surface 100-0 (head # 4) of -2 (the lowest error probability among the SSD0 and SSD2 selected in step S54).

In this embodiment, if the number of valid SSDs selected (detected) in step S55 is less than N (= 4), or if the number of valid SSDs is If they are not distributed on the disk surface, they are treated as device errors, and the processing described below is not performed.

When N valid SSDs (valid password information areas SSD) can be selected in step S55, the CPU 10 stores information on the N valid SSDs having position information of the N valid SSDs (valid SSDs). information)
Is registered in a predetermined area of the EEPROM 13 (step S
56).

As described above, if N valid SSDs distributed on each disk surface can be selected, and the information on the N valid SSDs (valid SSD information) can be correctly registered in a predetermined area of the EEPROM 13, CPU 10 is EE
The lock processing flag set in a predetermined area of the PROM 13 is reset (step S57), and the lock processing ends.

Therefore, when the lock processing flag is in the reset state at the end of the lock processing, it is guaranteed that the lock processing has been completed normally. If the power is shut off while writing password information or selecting a valid SSD, or if the lock process fails,
Since the process does not reach step S57, the in-process flag remains set.

Next, the lock release processing will be described with reference to the flowchart of FIG. First, when using the magnetic disk device of FIG. 1, the user performs an operation for releasing the locked state of the magnetic disk device from the host device. Thus, a command for unlocking (unlock command) including the password specified by the user is sent from the host device to the magnetic disk device of FIG.

The CPU 10 transmits the unlock command sent from the host device to the host interface 16 and the HD.
It is received via C14 (step S61). Then C
The PU 10 checks the state of the lock processing flag registered in a predetermined area of the EEPROM 13 (step S62). If the flag is set (step S63), the lock processing according to the flowchart of FIG. The lock release is aborted as it has been completed (step S64). In this case, the password is set again.

On the other hand, if the lock processing flag is not set (reset) (step 63), the CPU 10 determines that the lock processing according to the flowchart of FIG. 10 has been completed normally. In this case, the CPU 10 stores information (valid SSD information) on the valid SSD registered in a predetermined area of the EEPROM 13.
(Step S65), and based on the valid SSD information indicated by the valid SSD information, the disks 1-0 to 1-2 are read.
A total of N valid SSDs (valid password information areas S) distributed on N surfaces (here, N = 4) of the respective disk surfaces 100-0 and 100-1 (that is, a total of six disk surfaces).
SD), the disk read processing for sequentially reading the setting information (password information) is performed with the ECC collection operation ON (ON) (step S66).

Each time the password information is read from one valid SDD, the CPU 10 determines whether the read succeeded and whether the value of the read password information matches the checksum 45 attached to the password information. Is checked (step S67).

If the read fails or if the read succeeds but the checksum does not match, if the password information is not read from the last valid SSD (step S68), the CPU 10
The valid SSD information in is referred to again (step S65),
The password information is read from another valid SSD (step S66).

On the other hand, if the read is successful and the checksum is consistent, the CPU 10 sets the password 44 in the read password information and the user-specified password (added to the unlock command sent from the host device). (Step S69).
Only when they match (step S70), the locked state is released (step S71).

As described above, in the third password management control method, N valid SSDs are selected and registered each time a lock process (password setting process) is performed. When performing the processing, all of the N valid SSDs become available. In addition, the lock processing flag is set at the start of the lock processing, and the lock processing flag is reset when the lock processing ends normally. Therefore, the lock processing flag is reset to the reset state when the lock is released. If so, the password information of the N valid SSDs is guaranteed. Therefore, in order to prevent an adverse effect due to power interruption or the like at the time of setting password information, a counter unit is provided for password information as in the case where the first password management control method is applied, or the second password management control method is applied. It is not necessary to read the password information from the valid SSD in the same order as the writing of the password information as in the case where the password information is written.

However, in order to improve the access efficiency with respect to the password information, even when the first and third password management control methods are applied, the valid SSD information is provided with the order information and the N valid SSD information is provided. Or the order of access to
It is recommended that the order be such that access (seek) can be performed in the shortest time.

The above is one password information area SSD
Has been described as being composed of two sectors,
It may be composed of three or more sectors or one sector. If the password information area SSD
Is composed of one sector, it is not necessary to add the checksum 45 to the password information.

In the above description, the (four) password information areas SSD are provided on all the disk surfaces of the three disks.
The password information area SSD may be secured on at least two (plural) disk surfaces. The number of disks is not limited to three, but may be one. However, the greater the number of disk surfaces on which the password information areas SSD are dispersedly arranged, the higher the backup effect.

[0098]

As described above in detail, according to the present invention, even if the data in the password information setting area (password information area) secured on the disk is destroyed for some reason, another password information area is not used. The security function can be improved because backup is possible.

Further, according to the present invention, the password information can be efficiently accessed by providing only a predetermined number of password information areas having a low probability of occurrence of a read error as valid password information areas for password information access.

Further, according to the present invention, the password information write (password setting) and the password information read are performed in the same order, so that the integrity of the password information can be improved.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of a magnetic disk drive according to an embodiment of the present invention.

FIG. 2 shows a state in which four password information areas SSD0 to SSD3 are secured on both disk surfaces 100-0 and 100-1 of each disk 1-i (i = 0 to 2) in FIG. FIG.

FIG. 3 shows each password information area SSDj (j = 0 to 3).
FIG. 4 is a diagram showing an example of the sector configuration and a format example of password information set in the area SSDj.

FIG. 4 is a flowchart for explaining a valid SSD selection / registration process when the first password management control method is applied;

FIG. 5 is a flowchart for explaining a lock process (password setting process) when the first password management control method is applied.

FIG. 6 is a flowchart for explaining a lock release process when the first password management control method is applied.

FIG. 7 is a flowchart for explaining a valid SSD selection / registration process when the second password management control method is applied;

FIG. 8 is a flowchart for explaining lock processing (password setting processing) when the second password management control method is applied.

FIG. 9 is a flowchart for explaining a lock release process when the second password management control method is applied.

FIG. 10 is a flowchart for explaining a lock process (password setting process) when the third password management control method is applied.

FIG. 11 is a flowchart for explaining a lock release process when the third password management control method is applied.

[Explanation of symbols]

1-0 to 1-2, 1-i: disk, 2: head, 7: head IC, 8: read / write IC, 9: servo processing circuit, 10: CPU (password information writing means, valid password information area selection Means, registration means, determination means), 13 EEPROM (rewritable nonvolatile memory), 14 HDC (disk controller), 41 header section, 42 flag, 43 counter section, 44 password, 45 ... Checksum, 100-0, 100-1: Disk surface, 101: System area, SSD0 to SSD3, SSDi: Password information area, Sn, Sn + 1: Sector.

Continuation of the front page (58) Field surveyed (Int.Cl. ⁷ , DB name) G06F 3/06 G06F 1/00 G11B 20/10

Claims (7)

(57) [Claims] 1. A disk device having a security function in which a locked state for prohibiting disk access is released only when a password that matches a set password is given from a host device, wherein at least a recording medium a single disc,
A plurality of password information areas for recording password information have at least two disk surfaces out of all disk surfaces
At least one disk are distributed in the, from among all the password information area, a read error
Areas with a low probability of occurrence on at least two disk surfaces
Predetermined number of effective password information areas to be distributed
Selected by the effective password information area selecting means, and the effective password information area selecting means,
Predetermined distributed on at least two disk surfaces
Disk apparatus having a security function, characterized by comprising the password information writing means for writing the password <br/> over de information identical respectively to enable password information area number. 2. The method according to claim 1, wherein said valid password information area selecting means includes:
Of the specified number of valid password information areas
Recording means for recording the password information and, when accessing the password information,
Is determined based on the position information recorded in the recording means.
2. The disk device having a security function according to claim 1, further comprising means . 3. The method according to claim 2, wherein said valid password information area selecting means includes:
For a predetermined number of valid password information areas selected from
Recording means for recording information indicating the access order to be accessed , and the access order when accessing the password information.
Based on the access order information recorded in the recording means.
And means for determining.
2. A disk device having the security function according to 1 . 4. The method according to claim 1, wherein said valid password information area selecting means includes:
Of the specified number of valid password information areas
Location information and the predetermined number of valid password information areas.
Recording means for recording information indicating the access order to be accessed, and when accessing the password information, the access order.
Access the order in which they are recorded the access destination to the Symbol record means a
Information and means for determining based on the position information.
2. The disk device having a security function according to claim 1, wherein: Wherein the order indicated by the access order information, according to claim 3, characterized in that the probability of occurrence of a read error of the effective password information area of a predetermined number selected by the valid password information selecting means coincides low order Or a contract
A disk device having the security function according to claim 4 . 6. The order indicated by said access order information is
The predetermined number selected by the valid password information selecting means
Number of valid password information areas can be accessed in the shortest time
3. The method according to claim 3, wherein the values match in the proper order.
5. A disk device having the security function according to item 4 . 7. at least one disk as a recording medium, a plurality of paths for recording password information
The swward information area is at least 2 out of all disk surfaces
At least one disk that is distributed and
Comprising a disk, a password management method in a disk device having a security function was set to the locked state is released to prohibit the disk access only if the password matches the set password is given from the host device, Read error from all password information areas
Areas with a low probability of occurrence on at least two disk surfaces
Predetermined number of effective password information areas to be distributed
Selecting and distributing the selected at least two disk surfaces.
The specified number of valid password information areas
Writing the same password information respectively.
A password management method in a disk device having a security function.